A nondestructive screening method for retired power battery cells

Abstract

The application provides a lossless screening method for retired power battery cells, comprising the following steps: S1, obtaining the alternating current impedance spectrum data of the retired battery pack in a preset frequency range as a whole response signal, and separating a plurality of independent source signal components if the signal is non-Gaussian and independent of each other; S2, matching the separated source signal components with a standard impedance characteristic library established in advance to determine the cell type and aging mode corresponding to each source signal component; S3, for each identified cell, extracting the health state feature vector of each cell and inputting the health state feature vector of each cell into a pre-trained health state evaluation model; and S4, the health state evaluation model outputs the estimated values of the residual capacity and internal resistance of each cell, and if the residual capacity is higher than the availability threshold, it is judged that the cell has a value for step utilization. The application realizes lossless and accurate screening of retired battery cells, and significantly improves the resource utilization efficiency and economic benefits.

Description

Technical Field

[0001] This invention relates to the field of lithium battery recycling technology, and in particular to a non-destructive screening method for retired power battery cells. Background Technology

[0002] Lithium-ion battery recycling, as a crucial link in the sustainable development of the new energy industry, carries the dual mission of resource recycling and environmental protection. With the widespread adoption of electric vehicles, the number of retired power batteries has surged, making the efficient and environmentally friendly disposal of these batteries a key challenge for the industry. This field not only concerns economic benefits but also directly impacts the green transformation of the energy structure, urgently requiring innovative technologies to overcome existing bottlenecks.

[0003] Currently, battery condition assessment methods in lithium battery recycling often face problems of operational complexity and low efficiency. Many traditional methods require disassembling the battery pack and physically separating it to test the health of each internal cell. This approach is not only time-consuming and labor-intensive but can also cause irreversible damage to the battery structure and even increase safety hazards. More importantly, this method is difficult to adapt to the needs of large-scale industrial recycling, limiting the improvement of overall industry efficiency.

[0004] Against this backdrop, the technical challenges of accurately assessing the condition of cells within retired battery packs have become increasingly apparent. The primary issue is that, as a whole, the response signals from the cells within a battery pack often overlap, making it difficult to directly distinguish the true condition of individual cells. Furthermore, this signal mixing makes finding the specific frequency range most sensitive to cell health extremely difficult, because differences in the aging degree and internal characteristics of different cells interfere with the interpretation of the overall signal, leading to biased assessment results.

[0005] Therefore, how to separate and identify the unique response characteristics of each cell by applying external signals without disassembling the battery pack has become a key problem that urgently needs to be solved in the field of lithium battery recycling.

[0006] This problem is particularly prominent in actual business operations. For example, when dealing with a retired battery pack consisting of dozens of cells, technicians often cannot quickly determine which cells still have reuse value and which need to go directly into the dismantling process, leading to low efficiency in subsequent screening and resource allocation. This core challenge directly affects the overall process optimization and cost control of retired battery recycling. How to overcome the bottleneck of signal separation and accurate identification has become a key issue driving the development of the industry. Summary of the Invention

[0007] To address the technical problems mentioned in the background section, this invention provides a non-destructive screening method for retired power battery cells, the method comprising:

[0008] S1. Obtain the AC impedance spectrum data of the retired battery pack within a preset frequency range as the overall response signal. If the signal is non-Gaussian and independent, separate multiple independent source signal components. S2. Match the separated source signal components with a pre-established standard impedance feature library to determine the cell type and aging mode corresponding to each source signal component. S3. For each identified cell, extract the health status feature vector of each cell and input the health status feature vector of each cell into a pre-trained health status assessment model. S4. The health status assessment model outputs the estimated values ​​of the remaining capacity and internal resistance of each cell. If the remaining capacity is higher than the availability threshold, it is determined that the cell has cascade utilization value.

[0009] Furthermore, step S1 includes: applying a small-amplitude sinusoidal AC excitation to the battery using a dedicated testing device, and acquiring response signals at different frequencies.

[0010] Furthermore, the preset frequency range is 0.01 Hz to 10000 Hz.

[0011] Furthermore, a blind source separation algorithm is used to process the overall response signal.

[0012] Furthermore, step S2 includes: processing the original mixed impedance data using empirical mode decomposition technology to separate independent signal component units representing different frequency band characteristics.

[0013] Furthermore, the original mixed impedance data includes the superposition of ohmic internal resistance and charge transfer response, and independent high-frequency and mid-frequency component units are extracted by decomposition.

[0014] Furthermore, a dynamic time warping algorithm is used as the preset signal comparison method to compare the separated independent component units with the standard curves in the pre-constructed standard impedance library point by point, and calculate the similarity percentage as the impedance matching degree.

[0015] Furthermore, the characteristic peak values ​​of independent component units are extracted, and the aging mode category to which they belong is determined by the pattern classification method.

[0016] Furthermore, step S4 includes:

[0017] Step S41: Using the health status assessment model, obtain the remaining capacity data and internal resistance estimation data of each cell, remove outliers, and obtain a standardized cell status dataset.

[0018] Step S42: Extract the remaining capacity data and compare it with the preset availability threshold one by one. If the remaining capacity of a certain cell is higher than the preset availability threshold, it is determined that the cell initially meets the conditions for cascade utilization.

[0019] Step S43: Obtain the estimated internal resistance data of the battery cell that meets the conditions for tiered utilization, and perform a secondary analysis in combination with the internal resistance data. If the internal resistance data is within the preset range, the battery cell is further confirmed to have utilization value.

[0020] Step S44: Obtain valuable battery cell data, use a support vector machine model to classify and predict the health status of the battery cells, and obtain the classified battery cell status labels.

[0021] Step S45: Based on the classified cell status labels, obtain the tiered utilization potential level of each cell, and combine the remaining capacity and internal resistance data for comprehensive sorting to determine the priority utilization sequence of cells.

[0022] Step S46: For the priority battery cell sequence, obtain its detailed status data, generate a tiered utilization allocation scheme, and allocate the battery cells to the corresponding application scenarios through data matching technology to obtain the final utilization scheme.

[0023] The technical solution provided by this invention has the following beneficial effects:

[0024] This invention discloses a non-destructive screening method for retired power battery cells, proposing an innovative solution to the problem of assessing the health status of individual cells in retired battery packs and determining their value for secondary use. This invention acquires AC impedance spectrum data of the battery pack within a preset frequency range as the overall response signal, employs a blind source separation algorithm to separate independent source signal components, and matches them with a standard impedance feature library to accurately identify the cell type and aging mode. Subsequently, amplitude and phase change information in specific frequency bands are extracted as health status feature vectors, input into an evaluation model constructed based on a support vector regression algorithm, and the estimated remaining capacity and internal resistance of the cell are output. Finally, by comparing these values ​​with an availability threshold, the invention determines whether the cell has value for secondary use. This invention achieves non-destructive and accurate screening of retired battery cells, providing a scientific basis for secondary use and significantly improving resource utilization efficiency and economic benefits. Attached Figure Description

[0025] Figure 1 This is a flowchart of a non-destructive screening method for retired power battery cells according to the present invention.

[0026] Figure 2 This is a schematic diagram of a non-destructive screening method for retired power battery cells according to the present invention.

[0027] Figure 3 This is another schematic diagram of a non-destructive screening method for retired power battery cells according to the present invention. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0029] like Figures 1-3 As shown in this embodiment, a non-destructive screening method for retired power battery cells may specifically include:

[0030] S1: Obtain the AC impedance spectrum data of the retired battery pack within a preset frequency range as the overall response signal. If the signal is non-Gaussian and independent of each other, separate multiple independent source signal components.

[0031] Specifically, AC impedance spectral data of retired battery packs within a preset frequency range are acquired. Raw signals are collected using specialized testing equipment to form a comprehensive response signal data set. For this data set, a blind source separation algorithm is used for preliminary processing. If the signals meet the conditions of non-Gaussianity and mutual independence, multiple independent source signal components are separated. Frequency domain analysis is performed on the separated source signal components to extract the main frequency components of each component, obtaining the frequency domain characteristic distribution of each component. Based on the frequency domain characteristic distribution, the correspondence between each source signal component and different electrochemical processes within the retired battery pack is determined. If the frequency characteristics of a component match a preset electrochemical process frequency range, then that component is determined to be related to a specific process. Time domain analysis tools are used to further process the determined source signal components, obtaining the changing trends of each component over time, yielding time domain characteristic data. By comparing the time domain characteristic data with a preset battery degradation mode database, if the time domain characteristics of a component are consistent with the characteristic curve of a degradation mode in the database, then that component is determined to reflect a specific battery degradation mechanism. Based on the judgment results, each source signal component is correlated and mapped with the corresponding electrochemical process and degradation mechanism to obtain detailed decomposition information of the internal state of the retired battery pack.

[0032] Specifically, obtaining AC impedance spectrum data of retired battery packs within a preset frequency range is fundamental to assessing the internal state of the battery. A small-amplitude sinusoidal AC excitation is applied to the battery using specialized testing equipment, and response signals at different frequencies are collected.

[0033] For example, by setting the frequency range to 0.01 Hz to 10000 Hz, the device records a set of overall response signal data including phase differences and amplitude variations. This allows for non-destructive probing of complex electrochemical reactions within the battery, providing a rich data source for subsequent analysis.

[0034] It is understandable that the overall response signal is often a mixture of multiple internal reactions. Using a blind source separation algorithm, the mixed signal can be separated based solely on the statistical properties of the signals satisfying non-Gaussianity and mutual independence, even when the mixing mechanism of the source signals is unknown.

[0035] For example, the algorithm successfully separated three independent source signal components, each corresponding to a different physicochemical process within the battery. This processing effectively eliminates mutual interference between signals and significantly improves the accuracy of subsequent feature extraction.

[0036] Specifically, frequency domain analysis is performed on the separated source signal components to reveal their frequency characteristic distribution. The main frequency concentration regions of each component are extracted. If the frequency characteristics of the first component are found to be concentrated in the high-frequency region above 1000 Hz, according to electrochemical principles, this frequency range is usually related to ohmic internal resistance. If the second component is concentrated in the mid-frequency region from 1 Hz to 1000 Hz, it can be determined that it is related to charge transfer processes. If the third component is in the low-frequency region from 0.01 Hz to 1 Hz, it matches a solid-phase diffusion process. This establishes a precise correspondence between the abstract signal and the specific internal reactions of the battery.

[0037] Understandably, in order to investigate the battery degradation process, time-domain analysis tools are needed to process the determined source signal components and obtain their changing trends over time.

[0038] For example, the amplitude variation of the mid-frequency charge transfer component over the past 500 cycles is extracted to form time-domain feature data. This data is compared with a preset battery degradation mode database. If the amplitude of this component shows a specific exponential growth trend over time and is highly consistent with the characteristic curve of interface film thickening in the database, it can be determined that this component reflects this specific degradation mechanism.

[0039] Specifically, based on the above judgment results, each source signal component is deeply correlated and mapped with the corresponding electrochemical process and degradation mechanism.

[0040] For example, the system ultimately outputs a detailed breakdown, clearly indicating that the performance degradation of the retired battery pack is mainly due to the increased charge transfer impedance in the mid-frequency region, specifically manifested as thickening of the interface film. This multi-dimensional state decomposition not only improves the transparency of battery health status assessment but also provides reliable technical support for the secondary utilization of retired batteries.

[0041] S2, the separated source signal components are matched with a pre-established standard impedance characteristic library to determine the cell type and aging mode corresponding to each source signal component.

[0042] Specifically, source signal component data is acquired from the acquisition system, and signal decomposition technology is used to process the raw data, separating independent signal component units. Based on the separated signal component units, a preset signal comparison method is used to match them one by one with a standard impedance library to obtain the impedance matching degree between each component unit and the data in the library. For signal component units with high impedance matching degrees, if the matching degree exceeds a preset matching degree threshold, the cell type corresponding to that component is determined, and the type determination result is recorded. Based on the type determination result, the corresponding aging mode category data is obtained, and the signal component units are classified using a mode classification method to determine their respective aging mode categories. Based on the aging mode category and combined with the cell health data in the standard impedance library, a health assessment method is used to evaluate the signal component units, obtaining a cell health score. After obtaining the cell health score, combined with the aging degree value data, the aging degree value of each signal component unit is determined through comparative analysis, completing a comprehensive evaluation. Based on the comprehensive evaluation results, data integration technology is used to associate and store the cell type, aging mode category, health score, and aging degree value, generating structured data records.

[0043] In one possible implementation, empirical mode decomposition (EMD) is used to process the original mixed impedance data from the source signal component data acquired from the acquisition system, separating independent signal component units representing different frequency band characteristics.

[0044] For example, the raw mixed impedance data contains the superposition of ohmic internal resistance and charge transfer response. By decomposition, independent high-frequency and mid-frequency component units can be separated, providing clean data for subsequent comparisons.

[0045] For example, after obtaining the separated signal component units, the system calls a pre-built standard impedance library, which includes the standard impedance characteristics of various types of cells under healthy conditions and different aging levels. A dynamic time warping algorithm is used as the preset signal comparison method to compare the separated independent component units with the standard curves in the library point by point, calculating the similarity percentage as the impedance matching degree. When the impedance matching degree reaches a preset matching degree threshold of 90%, the cell type corresponding to that component can be determined to be a lithium iron phosphate cell, and the determination result is recorded.

[0046] It should be noted that after identifying the cell type, the system acquires the corresponding aging mode data. By extracting the characteristic peak values ​​of the independent component units of the signal, a mode classification method is used to determine the aging mode category to which it belongs, for example, identifying it as active lithium loss. Subsequently, combined with the cell health benchmark data of the corresponding aging mode in the standard impedance library, a weighted scoring health assessment method is used for quantitative evaluation, resulting in a cell health score of 80 points.

[0047] In one possible implementation, after obtaining a cell health score of 80 points, the system introduces aging level data for comparative analysis. If the absolute value of the real part of the impedance of this component increases by 15 milliohms, the aging level of that signal component is determined to be moderate aging, completing a comprehensive evaluation. Finally, data integration technology is used to associate and store the cell type, aging mode, 80-point health score, and moderate aging level value, generating structured data records and improving the efficiency of tracing the status of retired batteries.

[0048] S3. For each identified cell, extract the health status feature vector of each cell and input the health status feature vector of each cell into the pre-trained health status assessment model.

[0049] Specifically, for each battery cell, source signal data is acquired using signal acquisition equipment, focusing on signal performance within a specific frequency band to complete preliminary data collection and obtain the cell's raw signal dataset. From this raw signal dataset, amplitude and phase change data within the specific frequency band are extracted. Signal processing techniques are used to filter and normalize this data to determine the cell's signal feature set. Based on this feature set, the amplitude and phase change data are combined to form a feature vector for each cell, constructing a feature vector matrix that reflects the cell's health status. This feature vector matrix is ​​input into a pre-trained support vector regression model. The model's internal regression algorithm analyzes the feature vectors to determine the cell's health status level. If the health status level output by the support vector regression model is lower than a preset health status threshold, the cell's feature vector is extracted a second time to obtain more detailed amplitude and phase change data, resulting in a supplementary feature dataset. Using this supplementary feature dataset, combined with the output of the support vector regression model, a deep analysis of the cell's health status is performed to determine the final health status assessment result.

[0050] In one possible implementation, for the acquisition of cell source signal data, the system injects a small AC excitation of a specific frequency into the target cell using a high-precision signal acquisition device. The focus is on the mid-frequency range of 100 Hz to 1000 Hz, continuously acquiring the cell's response signal to this excitation, thereby forming a raw signal dataset containing rich electrochemical reaction information.

[0051] For example, after acquiring the original signal dataset, the system extracts the core parameters within that specific frequency band. Specifically, this involves separating the voltage response amplitude attenuation data and the phase offset angle data between current and voltage. Due to potential electromagnetic interference at the acquisition site, the system uses a low-pass filtering algorithm to smooth the data and remove abnormal spike noise. Subsequently, using a max-min normalization method, the absolute amplitude is uniformly mapped to a standard range of 0 to 1, establishing a standardized set of cell signal characteristics.

[0052] It should be noted that, in order to meet the input requirements of the machine learning model, the system sequentially concatenates the normalized amplitude data and phase change data. For example, the amplitude feature of 0.6 is combined with the phase feature of 45 degrees to construct the feature vector of a single cell. When processing multiple cells in a battery pack, these feature vectors are arranged in order to jointly construct a feature vector matrix reflecting the overall health status of the batch of cells.

[0053] In one possible implementation, the system inputs the constructed feature vector matrix into a pre-trained support vector regression model. This model maps the nonlinear features to a high-dimensional space using an internal kernel function, searches for the optimal regression hyperplane, and thus outputs a quantified health status level. Assume the model's predicted health status score for a specific battery cell is 75.

[0054] For example, the system's preset health status threshold is 80 points. When an output score of 75 is detected as being below this threshold, the system immediately triggers a secondary feature extraction mechanism. At this time, the acquisition device shifts its focus to the low-frequency band of 1 Hz to 10 Hz, extracting more detailed low-frequency phase delay data as a supplementary feature dataset. Finally, the system comprehensively analyzes the low-frequency supplementary data with the preliminary prediction results of the support vector regression model to determine that the true health status of the cell is severe degradation caused by localized lithium plating in the negative electrode material.

[0055] S4, the health status assessment model outputs the estimated values ​​of the remaining capacity and internal resistance of each cell, compares the estimated value of the remaining capacity of each cell with the preset availability threshold, and if the remaining capacity is higher than the availability threshold, it is determined that the cell has the value of secondary use.

[0056] Optionally, this step also includes:

[0057] Step S41: Using the health status assessment model, obtain the remaining capacity data and internal resistance estimation data of each cell, remove outliers, and obtain a standardized cell status dataset.

[0058] Step S42: Extract the remaining capacity data and compare it with the preset availability threshold one by one. If the remaining capacity of a certain cell is higher than the preset availability threshold, it is determined that the cell initially meets the conditions for cascade utilization.

[0059] Step S43: Obtain the estimated internal resistance data of the battery cell that meets the conditions for tiered utilization, and perform a secondary analysis in combination with the internal resistance data. If the internal resistance data is within the preset range, the battery cell is further confirmed to have utilization value.

[0060] Step S44: Obtain usable battery cell data, use a support vector machine model to classify and predict the health status of the battery cells, and obtain the classified battery cell status labels.

[0061] Step S45: Based on the classified cell status labels, obtain the tiered utilization potential level of each cell, and combine it with the remaining capacity and internal resistance data for comprehensive sorting to determine the priority utilization sequence of cells.

[0062] Step S46: For the priority battery cell sequence, obtain its detailed status data, generate a tiered utilization allocation scheme, and allocate the battery cells to the corresponding application scenarios through data matching technology to obtain the final utilization scheme.

[0063] Specifically, for the raw data of the collected battery cells, the estimated remaining capacity and internal resistance are first extracted. Due to potential environmental interference during the data collection process, outliers may exist in the data, deviating from the normal range. By setting reasonable upper and lower limits, these outliers are removed, and the remaining data is standardized to unify the data units.

[0064] For example, suppose a batch of battery cells has a rated capacity of 50 amp-hours. In the initial screening stage, a preset availability threshold for the remaining capacity is set to 35 amp-hours. The system compares each cell with the standardized dataset. If the current remaining capacity of a cell is 38 amp-hours, which is higher than the availability threshold of 35 amp-hours, then the cell is determined to initially meet the conditions for tiered utilization, thus forming the initial screening result.

[0065] In one possible implementation, based on the initial screening results, internal resistance data is further introduced for a second, in-depth analysis. Internal resistance is a key indicator reflecting the degree of aging within the battery cell.

[0066] For example, the preset range for internal resistance suitable for tiered utilization is set to 2.0 milliohms to 4.5 milliohms. If the estimated internal resistance of the aforementioned battery cell with a remaining capacity of 38 amp-hours is 3.2 milliohms, which falls within this preset range, then the battery cell is confirmed to have practical utilization value and is included in the secondary screening results. If the internal resistance is as high as 5.0 milliohms, it indicates severe internal aging and the cell is discarded.

[0067] Specifically, after acquiring usable battery cell data, a support vector machine model is used to classify and predict the battery cells. This model categorizes battery cells into different health status labels, such as good, moderate, and average, based on the input capacity and internal resistance characteristics.

[0068] For example, cells marked as good are assigned a higher potential level for reuse. Then, a comprehensive ranking is performed based on the remaining capacity and internal resistance, with cells having higher capacity and lower internal resistance ranked higher, thus determining the priority sequence of cells for reuse.

[0069] In one possible implementation, a specific tiered utilization allocation scheme is generated for the priority cell sequence. Through data matching technology, cells with different status tags are assigned to the most suitable application scenarios.

[0070] For example, cells with higher priority and good status are allocated to backup power supply scenarios for communication base stations with higher power requirements, while cells with average status and lower priority are allocated to energy storage scenarios for solar streetlights with lower charging and discharging requirements, thereby completing the final utilization scheme.

[0071] If the technical solution of this application involves the collection, processing, or application of personal information, the relevant products have, before implementing any personal information processing activities, fully and clearly informed individuals of the processing rules in accordance with the "Personal Information Protection Law of the People's Republic of China" and other current laws and regulations, and obtained their voluntary and explicit consent. If sensitive personal information is involved, the product has obtained the individual's separate consent before processing, and such consent is given in an explicit manner. For example, prominent signs are set up in the area where information collection devices such as cameras are located, clearly indicating "Entering is considered as consent to the collection of personal information"; or through pop-ups, checkboxes, user-initiated uploads, etc., under the premise of clearly listing the processor's identity, processing purpose, processing method, and information type, the user actively completes the authorization operation. The above mechanisms ensure that all personal information processing activities are based on legal authorization and fully comply with national compliance requirements regarding personal information protection.

[0072] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A non-destructive screening method for retired power battery cells, characterized in that, The method includes: S1. Obtain the AC impedance spectrum data of the retired battery pack within a preset frequency range as the overall response signal. If the signal is non-Gaussian and independent, separate multiple independent source signal components. S2. Match the separated source signal components with a pre-established standard impedance feature library to determine the cell type and aging mode corresponding to each source signal component. S3. For each identified cell, extract the health status feature vector of each cell and input the health status feature vector of each cell into a pre-trained health status assessment model. S4. The health status assessment model outputs the estimated values ​​of the remaining capacity and internal resistance of each cell. If the remaining capacity is higher than the availability threshold, it is determined that the cell has cascade utilization value.

2. The method according to claim 1, characterized in that, Step S1 includes: applying a small-amplitude sinusoidal AC excitation to the battery using a dedicated testing device, and collecting response signals at different frequencies.

3. The method according to claim 2, characterized in that, The preset frequency range is 0.01 Hz to 10000 Hz.

4. The method according to claim 3, characterized in that, A blind source separation algorithm is used to process the overall response signal.

5. The method according to claim 1, characterized in that, Step S2 includes: processing the original mixed impedance data using empirical mode decomposition technology to separate independent signal component units representing different frequency band characteristics.

6. The method according to claim 5, characterized in that, The original mixed impedance data contains the superposition of ohmic internal resistance and charge transfer response, and independent high-frequency and mid-frequency component units are extracted by decomposition.

7. The method according to claim 6, characterized in that, The dynamic time warping algorithm is used as the preset signal comparison method. The separated independent component units are compared with the standard curves in the pre-constructed standard impedance library point by point, and the similarity percentage is calculated as the impedance matching degree.

8. The method according to claim 7, characterized in that, The characteristic peak values ​​of independent component units are extracted, and the aging mode category to which they belong is determined by the pattern classification method.

9. The method according to claim 1, characterized in that, Step S4 includes: Step S41: Using the health status assessment model, obtain the remaining capacity data and internal resistance estimation data of each cell, remove outliers, and obtain a standardized cell status dataset. Step S42: Extract the remaining capacity data and compare it with the preset availability threshold one by one. If the remaining capacity of a certain cell is higher than the preset availability threshold, it is determined that the cell initially meets the conditions for cascade utilization. Step S43: Obtain the estimated internal resistance data of the battery cell that meets the conditions for tiered utilization, and perform a secondary analysis in combination with the internal resistance data. If the internal resistance data is within the preset range, the battery cell is further confirmed to have utilization value. Step S44: Obtain valuable battery cell data, use a support vector machine model to classify and predict the health status of the battery cells, and obtain the classified battery cell status labels. Step S45: Based on the classified cell status labels, obtain the tiered utilization potential level of each cell, and combine the remaining capacity and internal resistance data for comprehensive sorting to determine the priority utilization sequence of cells. Step S46: For the priority battery cell sequence, obtain its detailed status data, generate a tiered utilization allocation scheme, and allocate the battery cells to the corresponding application scenarios through data matching technology to obtain the final utilization scheme.

Images